Priority sampling and visual attention for self-driving car

Simple item page
dc.contributor.advisorMora Colque, Rensso Victor Hugo
dc.contributor.authorFlores Benites, Victor
dc.date.accessioned2023-09-27T13:57:42Z
dc.date.available2023-09-27T13:57:42Z
dc.date.issued2023
dc.description.abstractEnd-to-end methods facilitate the development of self-driving models by employing a single network that learns the human driving style from examples. However, these models face problems such as distributional shift, causal confusion, and high variance. To address these problems we propose two techniques. First, we propose the priority sampling algorithm, which biases a training sampling towards unknown observations for the model. Priority sampling employs a trade-off strategy that incentivizes the training algorithm to explore the whole dataset. Our results show a reduction of the error in the control signals in all the models studied. Moreover, we show evidence that our algorithm limits overtraining on noisy training samples. As a second approach, we propose a model based on the theory of visual attention (Bundesen, 1990) by which selecting relevant visual information to build an optimal environment representation. Our model employs two visual information selection mechanisms: spatial and feature-based attention. Spatial attention selects regions with visual encoding similar to contextual encoding, while feature-based attention selects features disentangled with useful information for routine driving. Furthermore, we encourage the model to recognize new sources of visual information by adding a bottom-up input. Results in the CoRL-2017 dataset (Dosovitskiy et al., 2017) show that our spatial attention mechanism recognizes regions relevant to the driving task. Our model builds disentangled features with low cosine similarity, but with high representation similarity. Finally, we report performance improvements over traditional end-to-end models.es_PE
dc.description.uriTesis de maestríaes_PE
dc.formatapplication/pdf
dc.identifier.other1079965
dc.identifier.urihttps://hdl.handle.net/20.500.12590/17744
dc.language.isoeng
dc.publisherUniversidad Católica San pablo
dc.publisher.countryPE
dc.rightsinfo:eu-repo/semantics/openAccess
dc.rights.urihttps://creativecommons.org/licenses/by-nc/4.0/
dc.subjectVisual attentiones_PE
dc.subjectSelf-drivinges_PE
dc.subjectNon-identically distributed data distribution es_PE
dc.subjectEnd-to-end methodses_PE
dc.subject.ocdehttp://purl.org/pe-repo/ocde/ford#1.02.01
dc.titlePriority sampling and visual attention for self-driving cares_PE
dc.typeinfo:eu-repo/semantics/masterThesis
dc.type.versioninfo:eu-repo/semantics/publishedVersion
renati.advisor.dni42846291
renati.advisor.orcidhttps://orcid.org/0000-0003-4734-8752
renati.author.dni71962886
renati.discipline611017
renati.jurorOchoa Luna, Jose Eduardo
renati.jurorCamara Chavez, Guillermo
renati.jurorChancán, Marvin
renati.levelhttps://purl.org/pe-repo/renati/level#maestro
renati.typehttps://purl.org/pe-repo/renati/type#tesis
thesis.degree.disciplineCiencia de la Computaciónes_PE
thesis.degree.grantorUniversidad Católica San Pablo. Departamento de Ciencia de la Computación
thesis.degree.levelMaestríaes_PE
thesis.degree.nameMaestro en Ciencia de la Computaciónes_PE
thesis.degree.programEscuela Profesional Ciencia de la Computaciónes_PE
Files
Original bundle
Now showing 1 - 4 of 4
Thumbnail Image
Name:
FLORES_BENITES_VIC_PRI.pdf
Size:
19.79 MB
Format:
Adobe Portable Document Format
Download
No Thumbnail Available
Name:
TURNITIN.pdf
Size:
25.56 MB
Format:
Adobe Portable Document Format
Download
No Thumbnail Available
Name:
AUTORIZACIÓN.pdf
Size:
321.16 KB
Format:
Adobe Portable Document Format
Download
No Thumbnail Available
Name:
ACTA.pdf
Size:
58.52 KB
Format:
Adobe Portable Document Format
Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed upon to submission
Description:
Download
Collections
Tesis - Maestría en Ciencias de la Computación